Loss function or error rate using logistical regression

Hi there,

I am trying to perform logistic regression using a test set and training set.

My loss rate should be simply

errs = logisticPred ~= yTest;
testErrRateLogistic = 100*sum(errs)/numel(errs);

However, I don't know how to get the y values obtained from a logistic regression classification in MATLAB? Nothing in http://uk.mathworks.com/help/stats/mnrfit.html tells you anything about how to obtain the predictions, only the p values and F statistics for the overall model.

I have a predictor matrix of 6 variables, and a categorical response matrix that has only 2 categories (binary classification).